Over time adherence to oral medication erodes in many schizophrenia patients, increasing relapse risk. Improving term medication adherence prevents relapse and may improve other outcomes. Until now, the only long-acting psychotic medications were high potency, first-generation antipsychotics that carry a significant side effect burden are perceived as the treatment of choice only for patients with documented histories of non-compliance and relapse second-generation anti-psychotic, risperidone, will soon be available in a long-acting injectable formulation, risperidone microspheres. We propose an eight-site randomized, open label trial to compare its effectiveness to second-generation oral anti-psychotics. Over the 5 year study period, 304 in- and out-patients will be randomized, treated for up 30 months and assessed by both remotely located, and local independent masked assessors in a number of critical domain. Details of the study rationale and methodology are provided in the individual site applications. The coordinating team will be led by John M. Kane, MD and includes Stephen R, Marder, MD and Nina R. Schooler, PhD, a group with extensive expertise in directing long-term clinical trials of schizophrenia. Each has led major studies of the older depot anti-psychotics, and they have collaborated successfully on multi-site studies. They will direct the fine-tuning of study design, oversee and coordinate recruitment, staff training, monitoring of rater-reliability and data quality at all sites. They will oversee data management, statistical analysis and reporting of results. A DSMB will be supported by grant. The coordination model includes conference calls to insure that specific groups, e.g. project coordinators, interact regularly. Face to face meetings of research teams will also be employed. A detailed study manual and site visits by coordinating team members will enhance cross-site consistency in protocol implementation. The site P.I.s represent talent and expertise, but none of them bas ever led a study of this scope. Participation in this study under the guidance the coordinating team will position them to lead future multi-site research in a field that sorely needs experienced investigators. Generalizability will be enhanced by the randomized but open-label design, which mirrors actual treatment more closely than does a double-blind design. The study is timely;the broadest population for study will be available when risperidone microspheres first comes to the market. Once clinicians have begun to prescribe this medication, patients who have received it will be excluded from the study and generalizability will decline. The study addresses questions that are unlikely to be studied by pharmaceutical industry sponsors and will generate unique information regarding long-term treatment options for schizophrenia.